How to load the base model in the fine-tuning task of KWS
Closed this issue · 3 comments
taorui-plus commented
I want to fine-tune the KWS task with a custom corpus based on the well-trained model. The language is English.
the well-trained model source:
git clone https://huggingface.co/yfyeung/icefall-asr-gigaspeech-zipformer-2023-10-17
project:
egs/gigaspeech/KWS
Execute the script:
./run.sh stage=3
Report an error:
Traceback (most recent call last):
File "/workspace/icefall/icefall/egs/gigaspeech/KWS/./zipformer/finetune.py", line 644, in <module>
main()
File "/workspace/icefall/icefall/egs/gigaspeech/KWS/./zipformer/finetune.py", line 638, in main
run(rank=0, world_size=1, args=args)
File "/workspace/icefall/icefall/egs/gigaspeech/KWS/./zipformer/finetune.py", line 464, in run
sp.load(params.bpe_model)
File "/opt/conda/lib/python3.10/site-packages/sentencepiece/__init__.py", line 961, in Load
return self.LoadFromFile(model_file)
File "/opt/conda/lib/python3.10/site-packages/sentencepiece/__init__.py", line 316, in LoadFromFile
return _sentencepiece.SentencePieceProcessor_LoadFromFile(self, arg)
RuntimeError: Internal: could not parse ModelProto from icefall-asr-gigaspeech-zipformer-2023-10-17/data/lang_bpe_500/bpe.model
I didn't find where the mistake was,can you help me? @pkufool @JinZr
JinZr commented
does this file exist icefall-asr-gigaspeech-zipformer-2023-10-17/data/lang_bpe_500/bpe.model
csukuangfj commented
Could you post the file size of the bpe.model
?
csukuangfj commented
Fixed in our wechat group by using
cd icefall-asr-gigaspeech-zipformer-2023-10-17/data/lang_bpe_500
git lfs pull --include "bpe.model"